INTERNATIONAL TELECOMMUNICATION UNION
AD HOC GROUP ON COST RECOVERY
FOR SATELLITE NETWORK FILINGS
Document INFO/4
23 February 2004
English only
GENEVA, 24 - 26 FEBRUARY 2004
Information Document by the Radiocommunication Bureau
METHODOLOGY FOR DETERMINING COST RECOVERY FEES
1. INTRODUCTION
Processing of satellite network filings, submitted under various provisions of Radio Regulation, was
described in detail by the BR in document INFO/1, submitted to the October 2003 meeting of the
Ad Hoc Group on Cost Recovery for Satellite Network Filings (AhGCR).
In document INFO/2, the BR described its time study, going on since the autumn of 2001, related
to various processes described in document INFO/1, and presented initial indications from the
analysis of data collected in the time study.
In the present document, the BR reports the results of further studies related to cost recovery for
satellite network filings and, drawing on the indications from those studies, suggests the
methodology for determination of cost recovery fees.
2. SUMMARY
Studies of time recorded for processing various satellite network filings indicate that in general the
time needed to process a filing contains a fixed time per filing and a variable time that depends on
the complexity and size of the filed network. This can be represented by a linear function as
follows:
y = A + Bx
where y is the total time for a type of filing, x is the size of the network in terms of cost recovery
units, A is the fixed (starting) time component and B is the time per unit. A and B are constants, the
line parameters, which are different for different processes (Advance Publication, Coordination,
Notification and Plans treatment) and for different complexities within each process. Constant B
may be zero, for cases for which the time needed does not depend on size.
The time study concentrated on developing relative y lines, i.e. relative constants A and B,
normalized to average time.
1/13
Once these relative constants are established, to calculate Cost Recovery Fees it is sufficient to
establish total cost for a process per year and total number of filings (cases) expected to be received
per year. Dividing the total cost by the number of filings gives the average cost. Multiplying the
relative y-line constants A and B by the average cost gives then the Cost Recovery Fee which
includes the starting fee (derived from A) and the fee per cost recovery unit (derived from B).
Table 1 provides the summary of data for calculation of the cost recovery fees. For each process, it
also includes the number of cases expected to be received per year, based on the Operational Plan of
the BR for 2004-2007, and the suggested maximum number of units to be taken into account for
calculation of cost recovery amounts for individual filings.
2/13
Table 1
Summary of data for calculation of Cost Recovery Fees
Type and Category
1
1
2
3
4
5
6
7
2
API (A)
Coordination (C)
3
A1
C1
C2
C3
Description
4
Article 9, IB
9.5, API/B
Article 9, IA
FoC1
FoC2,3
FoC4+
9.5D
CR/D
9.41/9.42
5
N4
Non-GSO
9
N5
Earth Stations
Resolution 49
Assistance SS
Assistance ES
Resubmissions
11.32A
11.41
11.47
11.49
Resolution 4
To be included in N
To be included in N
To be included in N
To be included in N
To be included in N
To be included in N
To be included in N
To be included in N
To be included in N
Ap30/30A
Ap30/30A
Ap30/30A
Ap30B
Ap30B
Article 4, Part A or B
Article 5
Article 2A
Article 6 or 7
Article 8
Plans (P)
7
Percentage of
Resources
(or of Total Cost)
8
Relative function (line)
A+Bx
A
B
11
12
Ratio to
Reference
Average
9
10
Maxim
number
of Units
13
1
0.86
1.03
2.00
0.7472
0.9032
1.8170
0.00002379
0.00002379
0.00002379
85
1
0.79
1.48
2.5 (?)
0.7204
1.433
2.0
0.0000683
0.0000683
0.0000683
25
3.5
1
1.00
1
0
-
160
11.5
1
1.00
1
0
-
102
35.5
10
23
10.5
57.8
9.6
4.6
20.4
7.6
1.00
0.40
0.80
1.00
0.80
0.946
0.3856
0.8
0.6509
0.52072
0.0002
0.0002
0
0.2222
0.2222
100
50 000
To be included in C
To be included in C
To be included in C
8
P1
P2
P3
P4
P5
6
25
250
G
10
11
12
13
14
Cases per year
To be included in A
To be included in C
N1
N2
N3
Notification (N)
FoC1
FoC2,3
FoC4+
Comment
SS: GSO and N-GSO
under coordination
Not under coordination
or only under 9.21
80
24
1
105
3/13
25 000 ?
10 000
10
3. COORDINATION
In document INFO/2, the BR reported preliminary results of data analysis concerning coordination
suggesting four complexity levels. In fact, suggested complexity categories of FoC-2 and FoC-3
have very similar average and starting values and comparable slopes (see Figures 5 and 10 of
INFO/2). It was therefore decided to conduct further analysis by combining these two groups into
one, henceforth called FoC2,3.
The period between the October 2003 and February 2004 meetings of the AhGCR was too short to
significantly increase the number of coordination requests with time data for all tasks. Due to this,
each one of some 20 tasks was separately analyzed for the number of cases in which it appeared. To
make it possible to reasonably aggregate individual averages into one common (total) average, two
correction factors were introduced. The first (Salary Factor in table 2 below), took into account the
different levels of staff executing different tasks. Level P3 was taken as the reference. The second
(Correction Factor in Table 2 below), corrected unrealistically low average times for certain tasks.
This second correction has no significant impact on overall results. Corrected averages for
individual tasks were then summed up to produce overall average. This overall average was then
used as the reference to normalize all time data and produce normalized (relative) best-fit lines for
each tasks and for each complexity level. Finally, since all these lines are based on the common
reference, they were added up to obtain overall best-fit lines for cases of FoC1, FoC2,3 and FoC4+.
Results are presented in Table 2 below and in Figure 1. As is evident from Figure 1, slopes for
FoC1 and FoC2,3 are very similar. For simplicity, it is suggested to use a common average slope
for these two cases. The highest complexity level, FoC4+, exhibits a much higher slope. However,
the number of these cases and their total time (resembling total cost) is small compared with FoC1
and FoC2,3. To further simplify the cost recovery fee structure, it is also possible to use for FoC4+
case the same slop as for Foc1 and FoC2,3 by turning its line around the average. The added benefit
of this is that the resultant per unit fee would be the same for any complexity within coordination.
The result is shown in Figure 3 and is included in the summary of Table 1.
All results are based on the same common reference. With this, it is possible to calculate cost
recovery fees directly from overall average cost per coordination request.
4/13
Table 2
COORDINATION – Data for calculation of Cost Recovery Fees
y = A + Bx
FOC = 2 & 3
FOC = 1
Task No
1
2
3
4
5
6
7
8
9
10
11
13
14
15
16
17
18
19
20
21
22
24
Task
Registration
As-received publication
Pre-validation
Notice preparation
Data capture
GIMS preparation
GIMS capture
Validation
Pcom and Completeness
Assistance
Restructure of Notice
Examination 1
Examination 2
Prep. Doc. for approval
Examination control
Finding capture
Preparation of Publication
Verification
Final Verification
Update of database
Info. to adm. & corresp.
Change to DBiU
Number of
Correction
Salary Factor
Cases
Factor
315
0.71
2.25
279
0.71
4.66
290
0.84
238
0.84
590
0.64
308
0.9
200
0.71
924
0.93
902
1
4
1
0.93
1431
1.08
1419
1.08
1431
1.08
1422
1.25
309
0.64
2.19
269
0.74
1.38
84
0.91
1.67
114
0.81
1.79
12
0.8
58
0.88
80
0.71
Normalized
to Total Avg
0.01324
0.01986
0.06588
0.02821
0.01508
0.03891
0.02270
0.05008
0.01122
Total
FOC = 4+
A
B
A
B
A
B
1.761E-02
1.860E-02
5.811E-02
2.383E-02
9.947E-03
3.215E-02
2.574E-02
4.461E-02
1.140E-02
0
0
6.607E-06
2.814E-06
3.252E-06
2.091E-06
-6.340E-06
4.747E-06
0
1.186E-02
2.047E-02
5.275E-02
2.647E-02
1.472E-02
2.879E-02
1.376E-02
4.828E-02
1.116E-02
0
0
2.516E-06
4.937E-07
-1.605E-08
2.046E-06
1.441E-06
1.237E-08
0
9.466E-03
1.572E-02
1.172E-01
4.167E-02
6.525E-02
7.308E-02
6.138E-02
8.238E-02
1.159E-02
0
0
1.616E-06
-9.449E-06
-2.015E-07
1.122E-06
2.172E-06
-8.586E-07
0
9.169E-01
2.622E-05
6.297E-06
9.808E-06
8.519E-06
0.02752
0.03182
2.792E-02 -3.938E-08 2.646E-02
6.306E-02
1.291E-06 8.294E-02
6.098E-02
1.414E-06 3.114E-02
Tasks 19-20 taken together
Too few samples, not taken into account
1.686E-02
1.028E-06
2.506E-02
6.036E-06 -2.236E-01
3.004E-02
1.946E-06
3.331E-02 -9.005E-07
1.00000
7.472E-01
9.095E-05
Too few samples, not taken into account
0.20018
0.15593
0.07218
0.02991
0.02686
0.06211
0.06789
0.06043
5/13
3.214E-01
6.664E-06
4.646E-01
6.626E-06
Tasks 13-16 taken together
2.388E-02
5.155E-02
6.145E-02
4.939E-07
2.670E-06
6.855E-07
2.666E-05
9.032E-01
2.092E-05
1.312E+00
4.571E-05
All Tasks
Relative Total Time (to Total Average Time for All Tasks)
7.0
6.0
y = 9.095E-05x + 1.312E+00
FOC 4+
5.0
4.0
3.0
2.0
FOC 2&3
y = 2.092E-05x + 9.032E-01
1.0
FOC 1
y = 2.666E-05x + 7.472E-01
0.0
10 000
20 000
30 000
40 000
50 000
60 000
Units
Figure 1
Coordination: Best-fit lines resulting from the time study
All Tasks (Modified Slopes)
Relative Total Time (to Total Average Time for All Tasks)
7.0
6.0
5.0
4.0
FOC 4+
y = 2.379E-05x + 1.817E+00
3.0
FOC 2&3
y = 2.379E-05x + 9.032E-01
2.0
FOC 1
y = 2.379E-05x + 7.472E-01
1.0
0.0
10 000
20 000
30 000
40 000
Units
Figure 2
Coordination: Corrected Best-fit lines
6/13
50 000
60 000
4. NOTIFICATION
Notification process deals with three different groups of cases: a) notification of earth station
assignments, b) notification of assignments of Non-GSO stations that are not subject to
coordination (including those that are subject only to 9.21) and c) notification of assignments to
GSO space stations and to Non-GSO space stations that are subject to coordination (with the
exception of Non-GSOs under 9.21 which are included in b) above).
In INFO/2 it was indicated that within groups a) and b) above, called for short ES and n-GSO, there
is no much difference in complexity and size and that a flat for cost fee, based on average costs,
would be appropriate. Group c) above, henceforth called GSO, contains a mixture of less and more
complex and small and large cases. Further analysis was done only for GSO group.
For the purpose of assigning appropriate notification costs to ES, n-GSO and GSO groups, an
estimate of the share of processing time (which represents costs) was made as follows: 11.5% ES,
3.5% n-GSO and 85% GSO.
4.1 Notification of networks subject to coordination
In INFO/2, two possible methods of expressing complexity were studied, one based on the number
of applicable forms of coordination (FoC), similar to coordination, and the other based on the
number of coordination special sections covered by the notification and the similarity or difference
of notification with coordination data. They were showing similar trends. After additional analysis
were performed, it was found that changing the method of expressing complexity does not switch
significant number of notifications from low to high complexity category. That being so, it is
simpler to use for notification the same expression of complexity as for coordination, i.e. the
number of applicable forms of coordination (FoC). Further analysis was conducted an the same
manner as for coordination as explained in Section 3 above. The results are presented in Table 3.
The time study since October 2001 has generated only one case for FoC4+ and therefore no results
are indicated for this case in Table 3. Also, results for FoC2,3 exhibit negative dependence on size.
This is not logical and can be explained by very small sample size of some tasks where a single
point can dramatically change the slope or its direction and summing up of individual task lines
without weighting based on the number of samples.
Results, as they appear in Table 3, are also illustrated in Figure 3.
Inspecting data of Table 3 reveals that tasks 13-16 (examination phase) contributes about 70% to
the overall average and the slope of FoC2,3 case for tasks 13-16 is very close to the overall slope
for FoC1. Therefore, the slope for FoC1 might then also be used for FoC2,3, while retaining the
starting point of foC2,3. This is illustrated in Figure 4 and results are included in the summary of
Table 1. Also, for FoC4+ arbitrary start and average values were entered and the same slope as for
FoC1 and FoC2,3.
7/13
Table 3
NOTIFICATION – Data for calculation of Cost Recovery Fees for GSO + Non-GSO subject to coordination
y = A + Bx
FOC = 2 & 3
FOC = 1
Task No
1
2
3
4
5
6
7
8
9
11
13
14
15
16
17
18
19
20
22
Task
Registration
As-received publication
Pre-validation
Notice preparation
Data capture
GIMS preparation
GIMS capture
Validation
Pcom and Completeness
Restructure of Notice
Examination 1
Examination 2
Prep. Doc. For approval
Examination control
Finding capture
Preparation of Publication
Verification
Final Verification
Info. To adm. & corresp.
Number of
Correction
Salary Factor
Cases
Factor
44
0.71
3.20
40
0.71
5.71
22
0.62
2.91
21
0.62
2.12
78
0.64
1
1
1.50
1
0.71
6.00
33
0.94
2.62
32
0.88
22
0.9
131
1.1
127
1.1
131
1.03
144
1.1
107
0.7
7
1
1.40
20
0.94
2.22
6
0.94
39
0.88
Normalized
to Total Avg
0.00904
0.01356
0.02763
0.01777
0.00844
0.02865
0.01356
0.03591
0.01231
0.01563
0.21847
0.30609
0.08561
0.08844
0.02147
0.00955
0.01796
0.02045
0.04944
Total
1.00000
8/13
A
B
A
B
8.668E-03
1.254E-02
2.047E-02
9.887E-03
6.719E-03
0
0
4.386E-05
4.309E-05
-4.981E-08
1.054E-02
1.292E-02
5.470E-02
5.324E-02
1.478E-02
0
0
-2.516E-05
-6.497E-05
1.713E-06
2.833E-02
9.842E-03
1.009E-02
5.146E-01
-1.349E-05
0
-1.299E-07
3.764E-05
5.711E-02
1.677E-02
3.884E-02
1.048E+00
2.626E-05
0
-6.244E-06
6.070E-05
FOC = 4+
A
B
?
?
Tasks 13-16 taken together
1.281E-02
7.159E-03
1.409E-02
1.821E-02
4.697E-02
3.318E-06
-3.694E-06
-1.122E-05
-3.079E-05
-2.428E-07
2.928E-02
1.168E-05
4.042E-02
-1.080E-05
5.643E-02
-1.191E-06
7.204E-01
6.830E-05
1.433E+00
-1.005E-05
All Tasks
Relative Total Time (to Total Average Time for All Tasks)
7.0
6.0
5.0
FOC 1
y = 6.830E-05x + 7.204E-01
4.0
3.0
2.0
FOC 2&3
y = -1.005E-05x + 1.433E+00
1.0
0.0
10 000
20 000
30 000
40 000
50 000
60 000
Units
Figure 3
Notification: Best-fit lines for GSO resulting from the time study
All Tasks
Relative Total Time (to Total Average Time for All Tasks)
7.0
6.0
y = 6.830E-05x + 1.433E+00
Adjusted FOC 2&3
5.0
4.0
FOC 1
y = 6.830E-05x + 7.204E-01
3.0
2.0
1.0
0.0
10 000
20 000
30 000
40 000
50 000
Units
Figure 4
Notification: Corrected best-fit lines for GSO
9/13
60 000
5. PLANS
5.1 Appendices 30/30A, Article 4 - Part A Special Section - Alternative definition for the Cost
Recovery Units
The first goal of the additional studies was to find an alternative definition for the Cost Recovery
Unit that better represents the relationship between the time needed for processing of a notice and
its size. For this purpose the time sheets of 171 published Part A Special Sections were analyzed.
The results of this study are illustrated in the attached Figures 5 and 6. These figures show the
relative SNP staff time (compared to the average value) versus the number of units. Since Region 2
notices include both the downlink and feeder-link assignments, each such notice has been taken as
two separate notices with the time and units divided by 2. Figure 5 corresponds to the current
definition of the cost recovery units for App.30/30A and Figure 6 corresponds to an alternative. In
these Figures, the best line fit for predicting the relative time from the units is included.
The analysis indicated that the use of the alternative definition of the cost recovery unit increased
considerably the correlation factor and thus better represents the relationship between the relative
time and notice size (see correlation values in Figures 5 and 6).
In this alternative, the number of test points in the current definition was replace by the number of
earth station antenna types (having different diameter/reference patterns). This new definition of the
Cost Recovery Unit for Appendices 30/30A is: Product of the number of assigned frequencies ,
number of earth station antenna types, number of the designation of emissions, summed up
for all groups of frequency assignments.
This analysis resulted in the relative best fit line as follows: y = 0.946 + 0.0002x. This is included in
the summary of Table 1.
5.2 Appendices 30/30A, Article 4 - Part B Special Section
Due to the small number of the available time sheet samples for this category, analysis of the units
formula was not feasible for this category. However, the same formula of Part A would be
applicable to this category since notices corresponding to both categories have the same types of
characteristics.
The preliminary ratio between the average processing time for Part B and Part A that is based on
only 12 Part B captured time sheet, is about 1.5. The additional resource required for processing of
Part B corresponds to the establishment of coordination requirement based on the Part A publication
and the subsequent comments from administration to that publication. However, in accordance with
§ 4.1.10ter/4.2.14ter of Article 4 of Appendices 30/30A this task is now transferred to the Part A
publication. The differences between the required average resources for these two categories will
thus be substantially reduced. One option would be to include the Part B notices in the same
category as Part A and use a common value and formula for both types of notices. This is included
in the summary of Table 1.
5.3 Appendices 30/30A, Article 2A – Space Operation Functions
Currently, this type of notice is included under the existing Category 2 of cost recovery. The Bureau
does not yet have any data regarding processing time of such notices. However, investigation of
several captured notices shows that they are very small in terms of number of units. The regulatory
and technical examination of such notices involves different provisions of Articles 4 and 7 of
Appendices 30/30A and RR Article 9. The estimated ratio of SNP average resources for Article 2A
10/13
notice to that of Article 4 downlink plus feeder-link is about 0.4. The following options may be
considered for this type of notice:
-
Distribute evenly the total cost of Article 2A processing in the average cost of that of
Article 4; or
-
Consider a flat fee equal to 40% of the total average costs of that of Article 4 downlink
plus feeder-link. This option is included in the summary of Table 1.
5.4 Appendix 30B, Article 6 – Sections IB & II
Due to the small number of available time sheet samples for this category, the analysis of the units
formula was not feasible. The current BR filings show that a maximum value of 10 could be
consider for the number of units of notices submitted under this category. The current values of the
flat fee and additional fee due to excess units, might require adjustment after calculation of the
average cost per notice for this category.
Figure 7 illustrates results of the analysis of time sheets of 21 Article 6 processed notices. The
resulting relative best-fit line is: y = 0.6509 + 0.2222x. This is included in the summary of Table 1.
The use of the current definition of Cost Recovery Unit for Appendix 30B seems adequate. By
maximizing the number of chargeable units to 10, the maximum to minimum charge ration will be
less than 3.5, as illustrated in Figure 8.
5. Other activities relating to Plans that are not under the current cost recovery
5.5a) Appendices 30/30A, Article 5 (Notification)
A preliminary study of 21 Article 5 time sheets suggests that the ratio of average processing time of
an Article 5 to an Article 4-Part A notice is about 0.4, with the same slope. Therefore, the following
relative best fit line formula ca be used: y = 0.3784 + 0.0002x (included in the summary of Table 1).
5.5b) Appendix 30B, Article 6 – Sections I, IA & III, Article 7
The processing of notices submitted under the above mentioned provisions does not currently result
in publication of a Special Section. However, the required SNP resources for sections I, IA and III
are comparable to those of section II of Article 6 which is currently subject to cost recovery under
Category 9. Moreover, the estimated average per notice of the SNP resources for processing of
Article 7 is about twice that of Article 6.
5.5c) Appendix 30B, Article 8 (Notification)
The estimated ratio of the SNP average resource for Article 8 notice to that of Article 6 is about 0.8.
The same cost recovery formula adopted for Article 6 would be also applicable to Article 8 after
multiplying by the above mentioned ratio.
11/13
4.50
4.00
Relative Time
3.50
3.00
2.50
2.00
1.50
1.00
0.50
0.00
0
5000
10000
15000
20000
25000
30000
35000
40000
45000
50000
Units
y = 8E-06x + 0.9842
(171 samples)
Correlation:0.13
Figure 5
Plans: Ap.30/30A, Art.4, Part A, Relative times vs Units (current definition)
4.50
4.00
Relative Time
3.50
3.00
2.50
2.00
1.50
1.00
0.50
0.00
0
1000
y = 0.0002x + 0.964
2000
3000
4000
5000
6000
7000
8000
9000
Units (TP out - esAnt in)
(171 samples)
Correlation:0.25
Figure 6
Plans: Ap.30/30A, Art.4, Part A, Relative times vs Units
(new definition of the Unit – TP out esAnt in)
12/13
1.8
1.6
Relative Time
1.4
1.2
1.0
0.8
0.6
0.4
0.2
0.0
0
1
2
3
AP30B Cost Recv Units
y = 0.2222x + 0.6509
Correlation:0.25
Figure 7
Plans: Ap.30B, Relative times vs Units
3.5
3
Relative Cost
2.5
2
1.5
1
0.5
0
1
2
3
4
5
6
7
8
Units
Figure 8
Plans: Ap.30B, Relative best-fit line extended to 10 units
_____________
13/13
9
10